knfsd: add file to export stats about nfsd pools

Add /proc/fs/nfsd/pool_stats to export to userspace various
statistics about the operation of rpc server thread pools.

This patch is based on a forward-ported version of
knfsd-add-pool-thread-stats which has been shipping in the SGI
"Enhanced NFS" product since 2006 and which was previously
posted:

http://article.gmane.org/gmane.linux.nfs/10375

It has also been updated thus:

 * moved EXPORT_SYMBOL() to near the function it exports
 * made the new struct struct seq_operations const
 * used SEQ_START_TOKEN instead of ((void *)1)
 * merged fix from SGI PV 990526 "sunrpc: use dprintk instead of
   printk in svc_pool_stats_*()" by Harshula Jayasuriya.
 * merged fix from SGI PV 964001 "Crash reading pool_stats before
   nfsds are started".
Signed-off-by: Greg Banks <gnb@sgi.com>
Signed-off-by: Harshula Jayasuriya <harshula@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>

fs/nfsd/nfsctl.c

@@ -60,6 +60,7 @@ enum {
 	NFSD_FO_UnlockFS,
 	NFSD_Threads,
 	NFSD_Pool_Threads,
+	NFSD_Pool_Stats,
 	NFSD_Versions,
 	NFSD_Ports,
 	NFSD_MaxBlkSize,
@@ -172,6 +173,16 @@ static const struct file_operations exports_operations = {
 	.owner		= THIS_MODULE,
 };
 
+extern int nfsd_pool_stats_open(struct inode *inode, struct file *file);
+
+static struct file_operations pool_stats_operations = {
+	.open		= nfsd_pool_stats_open,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= seq_release,
+	.owner		= THIS_MODULE,
+};
+
 /*----------------------------------------------------------------------------*/
 /*
  * payload - write methods
@@ -1246,6 +1257,7 @@ static int nfsd_fill_super(struct super_block * sb, void * data, int silent)
 		[NFSD_Fh] = {"filehandle", &transaction_ops, S_IWUSR|S_IRUSR},
 		[NFSD_Threads] = {"threads", &transaction_ops, S_IWUSR|S_IRUSR},
 		[NFSD_Pool_Threads] = {"pool_threads", &transaction_ops, S_IWUSR|S_IRUSR},
+		[NFSD_Pool_Stats] = {"pool_stats", &pool_stats_operations, S_IRUGO},
 		[NFSD_Versions] = {"versions", &transaction_ops, S_IWUSR|S_IRUSR},
 		[NFSD_Ports] = {"portlist", &transaction_ops, S_IWUSR|S_IRUGO},
 		[NFSD_MaxBlkSize] = {"max_block_size", &transaction_ops, S_IWUSR|S_IRUGO},

fs/nfsd/nfssvc.c

@@ -546,3 +546,10 @@ nfsd_dispatch(struct svc_rqst *rqstp, __be32 *statp)
 	nfsd_cache_update(rqstp, proc->pc_cachetype, statp + 1);
 	return 1;
 }
+
+int nfsd_pool_stats_open(struct inode *inode, struct file *file)
+{
+	if (nfsd_serv == NULL)
+		return -ENODEV;
+	return svc_pool_stats_open(nfsd_serv, file);
+}

include/linux/sunrpc/svc.h

@@ -24,6 +24,15 @@
  */
 typedef int		(*svc_thread_fn)(void *);
 
+/* statistics for svc_pool structures */
+struct svc_pool_stats {
+	unsigned long	packets;
+	unsigned long	sockets_queued;
+	unsigned long	threads_woken;
+	unsigned long	overloads_avoided;
+	unsigned long	threads_timedout;
+};
+
 /*
  *
  * RPC service thread pool.
@@ -42,6 +51,7 @@ struct svc_pool {
 	unsigned int		sp_nrthreads;	/* # of threads in pool */
 	struct list_head	sp_all_threads;	/* all server threads */
 	int			sp_nwaking;	/* number of threads woken but not yet active */
+	struct svc_pool_stats	sp_stats;	/* statistics on pool operation */
 } ____cacheline_aligned_in_smp;
 
 /*
@@ -396,6 +406,7 @@ struct svc_serv *  svc_create_pooled(struct svc_program *, unsigned int,
 			sa_family_t, void (*shutdown)(struct svc_serv *),
 			svc_thread_fn, struct module *);
 int		   svc_set_num_threads(struct svc_serv *, struct svc_pool *, int);
+int		   svc_pool_stats_open(struct svc_serv *serv, struct file *file);
 void		   svc_destroy(struct svc_serv *);
 int		   svc_process(struct svc_rqst *);
 int		   svc_register(const struct svc_serv *, const unsigned short,

net/sunrpc/svc_xprt.c

@@ -318,6 +318,8 @@ void svc_xprt_enqueue(struct svc_xprt *xprt)
 		goto out_unlock;
 	}
 
+	pool->sp_stats.packets++;
+
 	/* Mark transport as busy. It will remain in this state until
 	 * the provider calls svc_xprt_received. We update XPT_BUSY
 	 * atomically because it also guards against trying to enqueue
@@ -355,6 +357,7 @@ void svc_xprt_enqueue(struct svc_xprt *xprt)
 	if (pool->sp_nwaking >= SVC_MAX_WAKING) {
 		/* too many threads are runnable and trying to wake up */
 		thread_avail = 0;
+		pool->sp_stats.overloads_avoided++;
 	}
 
 	if (thread_avail) {
@@ -374,11 +377,13 @@ void svc_xprt_enqueue(struct svc_xprt *xprt)
 		atomic_add(rqstp->rq_reserved, &xprt->xpt_reserved);
 		rqstp->rq_waking = 1;
 		pool->sp_nwaking++;
+		pool->sp_stats.threads_woken++;
 		BUG_ON(xprt->xpt_pool != pool);
 		wake_up(&rqstp->rq_wait);
 	} else {
 		dprintk("svc: transport %p put into queue\n", xprt);
 		list_add_tail(&xprt->xpt_ready, &pool->sp_sockets);
+		pool->sp_stats.sockets_queued++;
 		BUG_ON(xprt->xpt_pool != pool);
 	}
 
@@ -591,6 +596,7 @@ int svc_recv(struct svc_rqst *rqstp, long timeout)
 	int			pages;
 	struct xdr_buf		*arg;
 	DECLARE_WAITQUEUE(wait, current);
+	long			time_left;
 
 	dprintk("svc: server %p waiting for data (to = %ld)\n",
 		rqstp, timeout);
@@ -676,12 +682,14 @@ int svc_recv(struct svc_rqst *rqstp, long timeout)
 		add_wait_queue(&rqstp->rq_wait, &wait);
 		spin_unlock_bh(&pool->sp_lock);
 
-		schedule_timeout(timeout);
+		time_left = schedule_timeout(timeout);
 
 		try_to_freeze();
 
 		spin_lock_bh(&pool->sp_lock);
 		remove_wait_queue(&rqstp->rq_wait, &wait);
+		if (!time_left)
+			pool->sp_stats.threads_timedout++;
 
 		xprt = rqstp->rq_xprt;
 		if (!xprt) {
@@ -1114,3 +1122,93 @@ int svc_xprt_names(struct svc_serv *serv, char *buf, int buflen)
 	return totlen;
 }
 EXPORT_SYMBOL_GPL(svc_xprt_names);
+
+
+/*----------------------------------------------------------------------------*/
+
+static void *svc_pool_stats_start(struct seq_file *m, loff_t *pos)
+{
+	unsigned int pidx = (unsigned int)*pos;
+	struct svc_serv *serv = m->private;
+
+	dprintk("svc_pool_stats_start, *pidx=%u\n", pidx);
+
+	lock_kernel();
+	/* bump up the pseudo refcount while traversing */
+	svc_get(serv);
+	unlock_kernel();
+
+	if (!pidx)
+		return SEQ_START_TOKEN;
+	return (pidx > serv->sv_nrpools ? NULL : &serv->sv_pools[pidx-1]);
+}
+
+static void *svc_pool_stats_next(struct seq_file *m, void *p, loff_t *pos)
+{
+	struct svc_pool *pool = p;
+	struct svc_serv *serv = m->private;
+
+	dprintk("svc_pool_stats_next, *pos=%llu\n", *pos);
+
+	if (p == SEQ_START_TOKEN) {
+		pool = &serv->sv_pools[0];
+	} else {
+		unsigned int pidx = (pool - &serv->sv_pools[0]);
+		if (pidx < serv->sv_nrpools-1)
+			pool = &serv->sv_pools[pidx+1];
+		else
+			pool = NULL;
+	}
+	++*pos;
+	return pool;
+}
+
+static void svc_pool_stats_stop(struct seq_file *m, void *p)
+{
+	struct svc_serv *serv = m->private;
+
+	lock_kernel();
+	/* this function really, really should have been called svc_put() */
+	svc_destroy(serv);
+	unlock_kernel();
+}
+
+static int svc_pool_stats_show(struct seq_file *m, void *p)
+{
+	struct svc_pool *pool = p;
+
+	if (p == SEQ_START_TOKEN) {
+		seq_puts(m, "# pool packets-arrived sockets-enqueued threads-woken overloads-avoided threads-timedout\n");
+		return 0;
+	}
+
+	seq_printf(m, "%u %lu %lu %lu %lu %lu\n",
+		pool->sp_id,
+		pool->sp_stats.packets,
+		pool->sp_stats.sockets_queued,
+		pool->sp_stats.threads_woken,
+		pool->sp_stats.overloads_avoided,
+		pool->sp_stats.threads_timedout);
+
+	return 0;
+}
+
+static const struct seq_operations svc_pool_stats_seq_ops = {
+	.start	= svc_pool_stats_start,
+	.next	= svc_pool_stats_next,
+	.stop	= svc_pool_stats_stop,
+	.show	= svc_pool_stats_show,
+};
+
+int svc_pool_stats_open(struct svc_serv *serv, struct file *file)
+{
+	int err;
+
+	err = seq_open(file, &svc_pool_stats_seq_ops);
+	if (!err)
+		((struct seq_file *) file->private_data)->private = serv;
+	return err;
+}
+EXPORT_SYMBOL(svc_pool_stats_open);
+
+/*----------------------------------------------------------------------------*/